# Library configuaration
import pandas as pd
import numpy as np
import requests
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.express as px
import plotly.graph_objects as go
from matplotlib.patches import Patch
from matplotlib.gridspec import GridSpec
from scipy import stats
import geopandas as gpd
import warnings
from dotenv import load_dotenv 
import os
load_dotenv()
import plotly.io as pio
import plotly.graph_objects as go

# Set the default renderer for Plotly figures
pio.renderers.default = 'notebook'

warnings.filterwarnings('ignore')

# Jupyter inline plotting
%matplotlib inline

# Plot styling
plt.style.use('seaborn-v0_8-whitegrid')
sns.set_palette("Set2")
plt.rcParams['figure.dpi'] = 110
plt.rcParams['font.size'] = 10

print("✓ Libraries imported successfully!")

# Census API setup
API_KEY = os.getenv("CENSUS_API_KEY")
BASE_URL = "https://api.census.gov/data/2022/acs/acs5"

# Variables to collect
VARIABLES = {
    "geographic": ["NAME"],
    "labor_market": ["B23025_002E", "B23025_003E", "B23025_004E", "B23025_005E"],
    "education": ["B15003_021E", "B15003_022E"],
    "housing": ["B25003_001E", "B25003_002E", "B25070_007E", "B25070_008E", 
                "B25070_009E", "B25070_010E", "B25002_001E", "B25002_003E", "B19013_001E"],
    "connectivity": ["B28002_002E"],
    "transportation": ["B08301_010E"],
    "population": ["B01003_001E"]
}

# Flatten for API call
var_list = [var for category in VARIABLES.values() for var in category]
var_string = ",".join(var_list)

print(f"✓ Configured {len(var_list)} Census variables")

✓ Libraries imported successfully!
✓ Configured 19 Census variables

import requests

# Retrieve ACS tract-level data for all U.S. states
states = [
    '01', '02', '04', '05', '06', '08', '09', '10', '11', '12',
    '13', '15', '16', '17', '18', '19', '20', '21', '22', '23',
    '24', '25', '26', '27', '28', '29', '30', '31', '32', '33',
    '34', '35', '36', '37', '38', '39', '40', '41', '42', '44',
    '45', '46', '47', '48', '49', '50', '51', '53', '54', '55', '56'
]

# Collect data for all tracts nationwide
all_data = []

for state in states:
    params = {
        "get": var_string,
        "for": "tract:*",
        "in": f"state:{state}",
        "key": API_KEY,
    }

    response = requests.get(BASE_URL, params=params)

    if response.status_code == 200:
        rows = response.json()
        # Append header only once
        if not all_data:
            all_data.extend(rows)
        else:
            all_data.extend(rows[1:])  
    else:
        print(f"State {state}: request failed ({response.status_code})")

# Final dataset as list of rows
data = all_data

# Build DataFrame from API response
df = pd.DataFrame(data[1:], columns=data[0])

# Construct Census tract identifiers
df["tract_6"] = df["tract"].astype(str).str.zfill(6)
df["geoid"] = (
    df["state"].astype(str).str.zfill(2) +
    df["county"].astype(str).str.zfill(3) +
    df["tract_6"]
)

# Convert applicable columns to numeric
exclude = ["NAME", "state", "county", "tract", "tract_6", "geoid"]
numeric_cols = [c for c in df.columns if c not in exclude]

for col in numeric_cols:
    df[col] = pd.to_numeric(df[col], errors="coerce")

df.head()

# Load FFIEC 2022 income classification data
ffiec_path = "/Users/sofiahutton/Documents/Fall 2025 CMU Classes/visualizations with python /CensusTractList2022.xlsx"

# Read FFIEC tract sheet
ffiec = pd.read_excel(ffiec_path, sheet_name="2022 tracts")
ffiec.columns = ffiec.columns.str.strip()

# Build GEOID for merging
ffiec["geoid"] = ffiec["FIPS code"].astype(str).str.zfill(11)

# Identify income-related columns by partial match (handles naming variation)
mfi_col = [c for c in ffiec.columns if "mfi" in c.lower()][0]
pct_col = [c for c in ffiec.columns if "percentage" in c.lower()][0]
lvl_col = [c for c in ffiec.columns if "income level" in c.lower()][0]

# Standardize column names
ffiec_keep = ffiec[["geoid", mfi_col, pct_col, lvl_col]].rename(
    columns={
        mfi_col: "Tract MFI",
        pct_col: "Tract income percentage",
        lvl_col: "Tract income level",
    }
)

# Remove any existing FFIEC columns to avoid duplicate/suffixed columns on re-run
cols_to_remove = [
    "Tract MFI", "Tract income percentage", "Tract income level",
    "Tract MFI_x", "Tract MFI_y",
    "Tract income percentage_x", "Tract income percentage_y",
    "Tract income level_x", "Tract income level_y"
]
df = df.drop(columns=[c for c in df.columns if c in cols_to_remove], errors="ignore")

# Merge FFIEC indicators into the ACS dataset
df = df.merge(ffiec_keep, on="geoid", how="left")

# Helper function for formatted summary
def print_income_summary(label, subset):
    dist = subset["Tract income level"].value_counts()
    total = dist.sum()
    print(f"\n{label} Income Classification:")
    for level in ["Upper", "Middle", "Moderate", "Low", "Unknown"]:
        count = dist.get(level, 0)
        share = count / total if total > 0 else 0
        print(f"- {level:8s}: {count:5,} ({share:.1%})")

# Nation-level summary (all tracts)
print_income_summary("Nationwide", df)

# Pennsylvania-only summary (state FIPS 42)
df_pa = df[df["state"] == "42"]
print_income_summary("Pennsylvania (state = 42)", df_pa)

# Allegheny County summary (state 42, county 003)
df_allegheny = df[(df["state"] == "42") & (df["county"] == "003")]
print_income_summary("Allegheny County (state 42, county 003)", df_allegheny)

Nationwide Income Classification:
- Upper   : 22,302 (26.7%)
- Middle  : 34,720 (41.6%)
- Moderate: 18,811 (22.5%)
- Low     : 5,430 (6.5%)
- Unknown : 2,268 (2.7%)

Pennsylvania (state = 42) Income Classification:
- Upper   :   813 (23.6%)
- Middle  : 1,630 (47.3%)
- Moderate:   711 (20.6%)
- Low     :   206 (6.0%)
- Unknown :    86 (2.5%)

Allegheny County (state 42, county 003) Income Classification:
- Upper   :   115 (29.2%)
- Middle  :   137 (34.8%)
- Moderate:    83 (21.1%)
- Low     :    38 (9.6%)
- Unknown :    21 (5.3%)

# Core tract-level rates
df["employment_rate"]   = df["B23025_004E"] / df["B23025_003E"]
df["unemployment_rate"] = df["B23025_005E"] / df["B23025_003E"]
df["homeownership_rate"] = df["B25003_002E"] / df["B25003_001E"]
df["vacancy_rate"]       = df["B25002_003E"] / df["B25002_001E"]
df["broadband_rate"]     = df["B28002_002E"] / df["B01003_001E"]
df["transit_rate"]       = df["B08301_010E"] / df["B01003_001E"]

# Households paying ≥30% of income on rent
df["rent_burdened_count"] = (
    df["B25070_007E"] +
    df["B25070_008E"] +
    df["B25070_009E"] +
    df["B25070_010E"]
)

df.head()

# Column mapping
output_columns = {
    "tract_6": "Tract Code (6-digit)",
    "NAME": "Tract Name",
    "Tract MFI": "FFIEC Tract MFI (2022)",
    "Tract income percentage": "FFIEC Tract income % (2022)",
    "Tract income level": "FFIEC Tract income level (2022)",
    "B01003_001E": "Total Population",
    "B23025_003E": "Labor Force",
    "B23025_004E": "Employed",
    "B23025_005E": "Unemployed",
    "employment_rate": "Employment Rate",
    "unemployment_rate": "Unemployment Rate",
    "B15003_021E": "Associates Degree",
    "B15003_022E": "Bachelors or Higher",
    "B25003_001E": "Total Housing Units",
    "B25003_002E": "Owner-Occupied",
    "homeownership_rate": "Homeownership Rate",
    "B25002_003E": "Vacant Units",
    "vacancy_rate": "Vacancy Rate",
    "rent_burdened_count": "Rent Burdened (30%+)",
    "B19013_001E": "Median Household Income",
    "B28002_002E": "With Broadband",
    "broadband_rate": "Broadband Rate",
    "B08301_010E": "Public Transit Commuters",
    "transit_rate": "Transit Rate",
}

# Create viz-ready dataframe
df_viz = df.copy()
df_viz = df_viz.rename(columns=output_columns)

# Drop the B23025_002E column (duplicate of Labor Force)
if 'B23025_002E' in df_viz.columns:
    df_viz = df_viz.drop(columns=['B23025_002E'])
    print("✓ Dropped duplicate B23025_002E column")

# Clean up the Tract Name to show only the 6-digit code
df_viz['Tract Name'] = df_viz['Tract Code (6-digit)']

# Additional derived variables
df_viz['Distress_Category'] = df_viz['FFIEC Tract income level (2022)'].apply(
    lambda x: 'Low/Moderate Income' if x in ['Low', 'Moderate'] else 'Middle/Upper Income'
)
df_viz['Bachelors_Plus_Rate'] = (df_viz['Bachelors or Higher'] / df_viz['Total Population'] * 100)
df_viz['Rent_Burden_Rate'] = (df_viz['Rent Burdened (30%+)'] / df_viz['Total Housing Units'] * 100)

# Convert rates to percentages - From here on, all rate columns in df_viz are expressed as percentages (0–100), not proportions (0–1).
for col in ['Employment Rate', 'Unemployment Rate', 'Homeownership Rate', 
            'Vacancy Rate', 'Broadband Rate', 'Transit Rate']:
    df_viz[col] = df_viz[col] * 100

# Drop the individual rent burden components since we have the total
rent_burden_components = ['B25070_007E', 'B25070_008E', 'B25070_009E', 'B25070_010E', 'B25002_001E']
df_viz = df_viz.drop(columns=rent_burden_components)
print(f"✓ Dropped {len(rent_burden_components)} rent burden component columns")

print("✓ Visualization dataset ready")
print(f"Shape: {df_viz.shape}")
df_viz.head()

✓ Dropped duplicate B23025_002E column
✓ Dropped 5 rent burden component columns
✓ Visualization dataset ready
Shape: (84415, 31)

# Create geographic filters and CLEAN the data
allegheny = df_viz[(df_viz['state'] == '42') & (df_viz['county'] == '003')].copy()
pennsylvania = df_viz[df_viz['state'] == '42'].copy()
nationwide = df_viz.copy()

# Clean Median Household Income (remove negative values and extreme outliers)
for df_geo in [allegheny, pennsylvania, nationwide]:
    df_geo.loc[df_geo['Median Household Income'] <= 0, 'Median Household Income'] = np.nan
    df_geo.loc[df_geo['Median Household Income'] > 500000, 'Median Household Income'] = np.nan

key_metrics = ['Total Population', 'Employment Rate', 'Unemployment Rate',
               'Bachelors_Plus_Rate', 'Median Household Income', 'Broadband Rate',
               'Homeownership Rate', 'Vacancy Rate']

print("="*80)
print("SUMMARY STATISTICS: ALLEGHENY vs PENNSYLVANIA vs NATIONWIDE")
print("="*80)

# Create comparison table
summary_comparison = pd.DataFrame({
    'Allegheny County': [
        allegheny['Total Population'].sum(),
        allegheny['Employment Rate'].mean(),
        allegheny['Unemployment Rate'].mean(),
        allegheny['Bachelors_Plus_Rate'].mean(),
        allegheny['Median Household Income'].mean(),
        allegheny['Broadband Rate'].mean(),
        allegheny['Homeownership Rate'].mean(),
        allegheny['Vacancy Rate'].mean()
    ],
    'Pennsylvania': [
        pennsylvania['Total Population'].sum(),
        pennsylvania['Employment Rate'].mean(),
        pennsylvania['Unemployment Rate'].mean(),
        pennsylvania['Bachelors_Plus_Rate'].mean(),
        pennsylvania['Median Household Income'].mean(),
        pennsylvania['Broadband Rate'].mean(),
        pennsylvania['Homeownership Rate'].mean(),
        pennsylvania['Vacancy Rate'].mean()
    ],
    'Nationwide': [
        nationwide['Total Population'].sum(),
        nationwide['Employment Rate'].mean(),
        nationwide['Unemployment Rate'].mean(),
        nationwide['Bachelors_Plus_Rate'].mean(),
        nationwide['Median Household Income'].mean(),
        nationwide['Broadband Rate'].mean(),
        nationwide['Homeownership Rate'].mean(),
        nationwide['Vacancy Rate'].mean()
    ]
}, index=key_metrics)

print(f"\n{'Geography':<25} {'Allegheny':<20} {'Pennsylvania':<20} {'Nationwide':<20}")
print("-" * 90)
print(f"{'Census Tracts':<25} {len(allegheny):<20,} {len(pennsylvania):<20,} {len(nationwide):<20,}")
print("\nValues (Total Population = SUM, others = MEAN):")
print("-" * 90)

# Format each row nicely
for metric in key_metrics:
    allegheny_val = summary_comparison.loc[metric, 'Allegheny County']
    pa_val = summary_comparison.loc[metric, 'Pennsylvania']
    nation_val = summary_comparison.loc[metric, 'Nationwide']
    
    if metric == 'Total Population':
        print(f"{metric:<25} {allegheny_val:>20,.0f} {pa_val:>20,.0f} {nation_val:>20,.0f}")
    elif metric == 'Median Household Income':
        print(f"{metric:<25} ${allegheny_val:>19,.2f} ${pa_val:>19,.2f} ${nation_val:>19,.2f}")
    else:
        print(f"{metric:<25} {allegheny_val:>20.2f} {pa_val:>20.2f} {nation_val:>20.2f}")

# Data quality note
print("\n" + "="*90)
print("DATA QUALITY NOTE:")
print(f"  Allegheny: {allegheny['Median Household Income'].isna().sum()} tracts with missing/invalid income data")
print(f"  Pennsylvania: {pennsylvania['Median Household Income'].isna().sum()} tracts with missing/invalid income data")
print(f"  Nationwide: {nationwide['Median Household Income'].isna().sum()} tracts with missing/invalid income data")

================================================================================
SUMMARY STATISTICS: ALLEGHENY vs PENNSYLVANIA vs NATIONWIDE
================================================================================

Geography                 Allegheny            Pennsylvania         Nationwide          
------------------------------------------------------------------------------------------
Census Tracts             394                  3,446                84,415              

Values (Total Population = SUM, others = MEAN):
------------------------------------------------------------------------------------------
Total Population                     1,245,310           12,989,208          331,097,593
Employment Rate                          94.42                94.31                94.36
Unemployment Rate                         5.58                 5.69                 5.64
Bachelors_Plus_Rate                      17.60                13.92                14.07
Median Household Income   $          75,812.23 $          77,527.23 $          80,716.70
Broadband Rate                           39.81                35.45                34.06
Homeownership Rate                       62.99                68.40                64.77
Vacancy Rate                             10.20                 9.94                10.73

==========================================================================================
DATA QUALITY NOTE:
  Allegheny: 13 tracts with missing/invalid income data
  Pennsylvania: 63 tracts with missing/invalid income data
  Nationwide: 1517 tracts with missing/invalid income data

print("\n" + "="*90)
print("INCOME LEVEL BREAKDOWN: ALLEGHENY vs PENNSYLVANIA vs NATIONWIDE")
print("="*90)

# Create comparison for each geography
for geo_name, geo_data in [('Allegheny County', allegheny), 
                            ('Pennsylvania', pennsylvania), 
                            ('Nationwide', nationwide)]:
    print(f"\n{geo_name.upper()}")
    print("="*90)
    
    # Group by income level
    grouped = geo_data.groupby('FFIEC Tract income level (2022)')
    
    # Calculate: SUM for Total Population, MEAN for everything else
    income_breakdown = grouped.agg({
        'Total Population': 'sum',
        'Employment Rate': 'mean',
        'Unemployment Rate': 'mean',
        'Bachelors_Plus_Rate': 'mean',
        'Median Household Income': 'mean',
        'Broadband Rate': 'mean',
        'Homeownership Rate': 'mean',
        'Vacancy Rate': 'mean'
    })
    
    # Format and display
    print(f"\n{'Metric':<30} {'Low':<15} {'Moderate':<15} {'Middle':<15} {'Upper':<15}")
    print("-"*90)
    
    for metric in key_metrics:
        if metric == 'Total Population':
            print(f"{metric:<30}", end="")
            for income_level in ['Low', 'Moderate', 'Middle', 'Upper']:
                if income_level in income_breakdown.index:
                    val = income_breakdown.loc[income_level, metric]
                    print(f"{val:>14,.0f} ", end="")
                else:
                    print(f"{'N/A':>14} ", end="")
            print()
        elif metric == 'Median Household Income':
            print(f"{metric:<30}", end="")
            for income_level in ['Low', 'Moderate', 'Middle', 'Upper']:
                if income_level in income_breakdown.index:
                    val = income_breakdown.loc[income_level, metric]
                    print(f"${val:>13,.0f} ", end="")
                else:
                    print(f"{'N/A':>14} ", end="")
            print()
        else:
            print(f"{metric:<30}", end="")
            for income_level in ['Low', 'Moderate', 'Middle', 'Upper']:
                if income_level in income_breakdown.index:
                    val = income_breakdown.loc[income_level, metric]
                    print(f"{val:>14.2f} ", end="")
                else:
                    print(f"{'N/A':>14} ", end="")
            print()
    
    # Sample sizes
    print("\n" + "-"*90)
    print("Sample Sizes (number of census tracts):")
    tract_counts = geo_data['FFIEC Tract income level (2022)'].value_counts()
    print(f"{'Low:':<15} {tract_counts.get('Low', 0):>6,}  |  ", end="")
    print(f"{'Moderate:':<15} {tract_counts.get('Moderate', 0):>6,}  |  ", end="")
    print(f"{'Middle:':<15} {tract_counts.get('Middle', 0):>6,}  |  ", end="")
    print(f"{'Upper:':<15} {tract_counts.get('Upper', 0):>6,}")

==========================================================================================
INCOME LEVEL BREAKDOWN: ALLEGHENY vs PENNSYLVANIA vs NATIONWIDE
==========================================================================================

ALLEGHENY COUNTY
==========================================================================================

Metric                         Low             Moderate        Middle          Upper          
------------------------------------------------------------------------------------------
Total Population                      91,223        219,912        437,560        479,547 
Employment Rate                        89.25          92.55          95.30          96.72 
Unemployment Rate                      10.75           7.45           4.70           3.28 
Bachelors_Plus_Rate                     7.51          12.87          18.87          23.26 
Median Household Income       $       33,979 $       49,919 $       72,018 $      114,459 
Broadband Rate                         33.57          40.04          41.72          39.32 
Homeownership Rate                     34.57          55.04          67.35          75.52 
Vacancy Rate                           19.74          13.67           8.01           6.46 

------------------------------------------------------------------------------------------
Sample Sizes (number of census tracts):
Low:                38  |  Moderate:           83  |  Middle:            137  |  Upper:             115

PENNSYLVANIA
==========================================================================================

Metric                         Low             Moderate        Middle          Upper          
------------------------------------------------------------------------------------------
Total Population                     683,935      2,456,728      6,341,164      3,375,388 
Employment Rate                        88.66          92.35          95.12          95.95 
Unemployment Rate                      11.34           7.65           4.88           4.05 
Bachelors_Plus_Rate                     6.34           9.92          13.44          20.70 
Median Household Income       $       36,741 $       54,639 $       75,394 $      113,158 
Broadband Rate                         31.73          34.92          35.59          37.41 
Homeownership Rate                     36.68          56.63          74.12          76.89 
Vacancy Rate                           13.83          12.39           9.94           6.38 

------------------------------------------------------------------------------------------
Sample Sizes (number of census tracts):
Low:               206  |  Moderate:          711  |  Middle:          1,630  |  Upper:             813

NATIONWIDE
==========================================================================================

Metric                         Low             Moderate        Middle          Upper          
------------------------------------------------------------------------------------------
Total Population                  18,412,899     71,654,894    138,968,321     95,017,217 
Employment Rate                        89.52          92.95          95.01          95.91 
Unemployment Rate                      10.48           7.05           4.99           4.09 
Bachelors_Plus_Rate                     6.70           9.57          13.46          20.76 
Median Household Income       $       39,011 $       56,105 $       76,431 $      118,666 
Broadband Rate                         30.17          32.38          34.42          36.22 
Homeownership Rate                     31.72          52.88          70.10          76.12 
Vacancy Rate                           13.30          11.47          11.00           8.76 

------------------------------------------------------------------------------------------
Sample Sizes (number of census tracts):
Low:             5,430  |  Moderate:       18,811  |  Middle:         34,720  |  Upper:          22,302

print("\n" + "="*90)
print("OPPORTUNITY GAPS: LOW/MODERATE vs MIDDLE/UPPER INCOME")
print("Comparison across Allegheny County, Pennsylvania, and Nationwide")
print("="*90)

gap_metrics = ['Employment Rate', 'Unemployment Rate', 'Broadband Rate', 
               'Homeownership Rate', 'Vacancy Rate', 'Transit Rate']

# Calculate gaps for each geography
results = []

for geo_name, geo_data in [('Allegheny', allegheny), 
                            ('Pennsylvania', pennsylvania), 
                            ('Nationwide', nationwide)]:
    
    low_mod = geo_data[geo_data['Distress_Category'] == 'Low/Moderate Income'][gap_metrics].mean()
    mid_upper = geo_data[geo_data['Distress_Category'] == 'Middle/Upper Income'][gap_metrics].mean()
    gap = mid_upper - low_mod
    
    # Get population and tract counts
    low_mod_pop = geo_data[geo_data['Distress_Category'] == 'Low/Moderate Income']['Total Population'].sum()
    mid_upper_pop = geo_data[geo_data['Distress_Category'] == 'Middle/Upper Income']['Total Population'].sum()
    low_mod_tracts = len(geo_data[geo_data['Distress_Category'] == 'Low/Moderate Income'])
    mid_upper_tracts = len(geo_data[geo_data['Distress_Category'] == 'Middle/Upper Income'])
    
    results.append({
        'Geography': geo_name,
        'Low/Mod_n': low_mod_tracts,
        'Mid/Upper_n': mid_upper_tracts,
        'Low/Mod_pop': low_mod_pop,
        'Mid/Upper_pop': mid_upper_pop,
        **{f'{metric}_LowMod': low_mod[metric] for metric in gap_metrics},
        **{f'{metric}_MidUpper': mid_upper[metric] for metric in gap_metrics},
        **{f'{metric}_Gap': gap[metric] for metric in gap_metrics}
    })

# Display in organized format
for metric in gap_metrics:
    print(f"\n{metric.upper()}")
    print("-"*90)
    print(f"{'Geography':<15} {'Low/Moderate':<15} {'Middle/Upper':<15} {'Gap':<15} {'% Difference':<15}")
    print("-"*90)
    
    for result in results:
        low_mod_val = result[f'{metric}_LowMod']
        mid_upper_val = result[f'{metric}_MidUpper']
        gap_val = result[f'{metric}_Gap']
        pct_diff = (gap_val / low_mod_val * 100) if low_mod_val != 0 else 0
        
        print(f"{result['Geography']:<15} {low_mod_val:>14.2f} {mid_upper_val:>14.2f} {gap_val:>14.2f} {pct_diff:>14.1f}%")

# Sample sizes with population totals
print("\n" + "="*90)
print("SAMPLE SIZES & POPULATIONS")
print("="*90)
print(f"{'Geography':<15} {'Category':<20} {'Tracts':<15} {'Total Population':<20}")
print("-"*90)

for result in results:
    print(f"{result['Geography']:<15} {'Low/Moderate Income':<20} {result['Low/Mod_n']:>14,} {result['Low/Mod_pop']:>19,.0f}")
    print(f"{'':15} {'Middle/Upper Income':<20} {result['Mid/Upper_n']:>14,} {result['Mid/Upper_pop']:>19,.0f}")
    print()

==========================================================================================
OPPORTUNITY GAPS: LOW/MODERATE vs MIDDLE/UPPER INCOME
Comparison across Allegheny County, Pennsylvania, and Nationwide
==========================================================================================

EMPLOYMENT RATE
------------------------------------------------------------------------------------------
Geography       Low/Moderate    Middle/Upper    Gap             % Difference   
------------------------------------------------------------------------------------------
Allegheny                91.51          95.76           4.25            4.6%
Pennsylvania             91.52          95.34           3.81            4.2%
Nationwide               92.18          95.25           3.07            3.3%

UNEMPLOYMENT RATE
------------------------------------------------------------------------------------------
Geography       Low/Moderate    Middle/Upper    Gap             % Difference   
------------------------------------------------------------------------------------------
Allegheny                 8.49           4.24          -4.25          -50.0%
Pennsylvania              8.48           4.66          -3.81          -45.0%
Nationwide                7.82           4.75          -3.07          -39.3%

BROADBAND RATE
------------------------------------------------------------------------------------------
Geography       Low/Moderate    Middle/Upper    Gap             % Difference   
------------------------------------------------------------------------------------------
Allegheny                38.00          40.64           2.64            6.9%
Pennsylvania             34.20          35.91           1.71            5.0%
Nationwide               31.89          34.95           3.06            9.6%

HOMEOWNERSHIP RATE
------------------------------------------------------------------------------------------
Geography       Low/Moderate    Middle/Upper    Gap             % Difference   
------------------------------------------------------------------------------------------
Allegheny                48.61          69.63          21.01           43.2%
Pennsylvania             52.15          74.40          22.25           42.7%
Nationwide               48.14          71.59          23.45           48.7%

VACANCY RATE
------------------------------------------------------------------------------------------
Geography       Low/Moderate    Middle/Upper    Gap             % Difference   
------------------------------------------------------------------------------------------
Allegheny                15.58           7.72          -7.86          -50.5%
Pennsylvania             12.71           8.91          -3.80          -29.9%
Nationwide               11.88          10.26          -1.62          -13.6%

TRANSIT RATE
------------------------------------------------------------------------------------------
Geography       Low/Moderate    Middle/Upper    Gap             % Difference   
------------------------------------------------------------------------------------------
Allegheny                 6.02           3.13          -2.89          -48.1%
Pennsylvania              3.49           1.67          -1.82          -52.1%
Nationwide                2.64           1.54          -1.10          -41.7%

==========================================================================================
SAMPLE SIZES & POPULATIONS
==========================================================================================
Geography       Category             Tracts          Total Population    
------------------------------------------------------------------------------------------
Allegheny       Low/Moderate Income             121             311,135
                Middle/Upper Income             273             934,175

Pennsylvania    Low/Moderate Income             917           3,140,663
                Middle/Upper Income           2,529           9,848,545

Nationwide      Low/Moderate Income          24,241          90,067,793
                Middle/Upper Income          60,174         241,029,800

# 6.1 Opportunity gaps across income levels

import plotly.graph_objects as go
import plotly.express as px  # kept in case used elsewhere

# Metrics to compare between low/moderate- and middle/upper-income tracts
gap_metrics = [
    "Employment Rate",
    "Unemployment Rate",
    "Broadband Rate",
    "Homeownership Rate",
    "Vacancy Rate",
    "Bachelors_Plus_Rate",
]

gap_data = []

# Compute percentage differences by geography
for geo_name, geo_data in [
    ("Allegheny County", allegheny),
    ("Pennsylvania", pennsylvania),
    ("Nationwide", nationwide),
]:
    low_mod = geo_data[geo_data["Distress_Category"] == "Low/Moderate Income"][gap_metrics].mean()
    mid_upper = geo_data[geo_data["Distress_Category"] == "Middle/Upper Income"][gap_metrics].mean()

    for metric in gap_metrics:
        gap = mid_upper[metric] - low_mod[metric]
        pct_diff = (gap / low_mod[metric] * 100) if low_mod[metric] != 0 else 0

        gap_data.append(
            {
                "Geography": geo_name,
                "Metric": metric,
                "Gap": gap,
                "Percent_Diff": pct_diff,
            }
        )

gap_df = pd.DataFrame(gap_data)

# Colors by geography
colors = {
    "Allegheny County": "#1f77b4",  # blue
    "Pennsylvania": "#9467bd",      # purple
    "Nationwide": "#ff7f0e",        # orange
}

fig = go.Figure()

for geo in ["Allegheny County", "Pennsylvania", "Nationwide"]:
    geo_data = gap_df[gap_df["Geography"] == geo]

    fig.add_trace(
        go.Bar(
            y=geo_data["Metric"],
            x=geo_data["Percent_Diff"],
            name=geo,
            orientation="h",
            marker=dict(color=colors[geo], opacity=0.8),
            text=geo_data["Percent_Diff"].round(1).astype(str) + "%",
            textposition="outside",
            textfont=dict(size=11, family="Arial, sans-serif"),
            hovertemplate=(
                "<b>%{y}</b><br>"
                f"<b>{geo}</b><br>"
                "Gap: %{x:.1f}%<br>"
                "<extra></extra>"
            ),
        )
    )

fig.update_layout(
    title=dict(
        text="Opportunity Gaps Across Income Levels",
        x=0.5,
        xanchor="center",
        font=dict(size=24, family="Arial, sans-serif", color="#2c3e50"),
    ),
    annotations=[
        dict(
            text="Percentage difference: Middle/Upper Income minus Low/Moderate Income tracts",
            x=0.5,
            y=-0.15,
            xref="paper",
            yref="paper",
            xanchor="center",
            yanchor="top",
            showarrow=False,
            font=dict(size=12, color="#7f8c8d", family="Arial, sans-serif"),
        )
    ],
    xaxis=dict(
        title="Percentage Point Difference (%)",
        title_font=dict(size=14, family="Arial, sans-serif"),
        gridcolor="#ecf0f1",
        tickfont=dict(size=12, family="Arial, sans-serif"),
    ),
    yaxis=dict(
        title="",
        tickfont=dict(size=13, family="Arial, sans-serif"),
    ),
    barmode="group",
    height=550,
    width=1100,
    template="plotly_white",
    plot_bgcolor="white",
    paper_bgcolor="white",
    legend=dict(
        orientation="h",
        yanchor="top",
        y=1.12,
        xanchor="center",
        x=0.5,
        font=dict(size=13, family="Arial, sans-serif"),
        bgcolor="rgba(255,255,255,0.8)",
        bordercolor="#bdc3c7",
        borderwidth=1,
    ),
    font=dict(family="Arial, sans-serif"),
    margin=dict(t=120, b=100, l=150, r=80),
)

fig.show()

# 6.2 Economic opportunity scorecard heatmap

import plotly.graph_objects as go
import numpy as np

# Metrics to include in the heatmap
metrics_for_heatmap = [
    "Employment Rate",
    "Unemployment Rate",
    "Bachelors_Plus_Rate",
    "Median Household Income",
    "Broadband Rate",
    "Homeownership Rate",
    "Vacancy Rate",
]

heatmap_data = []

# Aggregate mean values by geography and FFIEC income level
for geo_name, geo_data in [
    ("Allegheny County", allegheny),
    ("Pennsylvania", pennsylvania),
    ("Nationwide", nationwide),
]:
    for income_level in ["Low", "Moderate", "Middle", "Upper"]:
        subset = geo_data[geo_data["FFIEC Tract income level (2022)"] == income_level]

        if len(subset) == 0:
            continue

        row_data = {
            "Geography": geo_name,
            "Income Level": income_level,
        }

        for metric in metrics_for_heatmap:
            if metric == "Median Household Income":
                # Convert to thousands for display
                row_data[metric] = subset[metric].mean() / 1000
            else:
                row_data[metric] = subset[metric].mean()

        heatmap_data.append(row_data)

heatmap_df = pd.DataFrame(heatmap_data)

# Row labels (geography + income tier)
heatmap_df["Label"] = heatmap_df["Geography"] + " - " + heatmap_df["Income Level"]

# Normalize all metrics to a 0–100 scale where higher = better outcome
z_data_normalized = []
display_values = []

for metric in metrics_for_heatmap:
    col_data = heatmap_df[metric].values
    min_val, max_val = col_data.min(), col_data.max()

    if max_val == min_val:
        # Avoid divide-by-zero; flat metric across all groups
        normalized = np.full_like(col_data, 50, dtype=float)
    else:
        if metric in ["Unemployment Rate", "Vacancy Rate"]:
            # Lower is better: invert so high normalized = better outcome
            normalized = 100 - ((col_data - min_val) / (max_val - min_val) * 100)
        else:
            # Higher is better
            normalized = (col_data - min_val) / (max_val - min_val) * 100

    z_data_normalized.append(normalized)
    display_values.append(col_data)

z_data_normalized = np.array(z_data_normalized).T  # shape: (rows, metrics)
display_values = np.array(display_values).T

# Axis labels
y_labels = heatmap_df["Label"].values
x_labels = [
    "Employment<br>Rate (%)",
    "Unemployment<br>Rate (%)",
    "Bachelor's<br>Degree+ (%)",
    "Median HH<br>Income ($K)",
    "Broadband<br>Access (%)",
    "Homeownership<br>Rate (%)",
    "Vacancy<br>Rate (%)",
]

# Custom hover text with original values
hover_text = []
for i, row in heatmap_df.iterrows():
    row_hover = []
    for j, metric in enumerate(metrics_for_heatmap):
        val = display_values[i][j]
        if metric == "Median Household Income":
            formatted_val = f"${val:.1f}K"
        else:
            formatted_val = f"{val:.1f}%"
        row_hover.append(
            f"<b>{metric}</b><br>{formatted_val}<br><b>{row['Label']}</b>"
        )
    hover_text.append(row_hover)

# Heatmap: colors reflect normalized scores, text shows original metric values
fig = go.Figure(
    data=go.Heatmap(
        z=z_data_normalized,
        x=x_labels,
        y=y_labels,
        colorscale="RdYlBu_r",  # red = weaker outcomes, blue = stronger
        text=np.round(display_values, 1),
        texttemplate="%{text}",
        textfont={"size": 10},
        hovertext=hover_text,
        hoverinfo="text",
        colorbar=dict(
            title="Outcome<br>Quality",
            titleside="right",
            tickmode="array",
            tickvals=[0, 50, 100],
            ticktext=["Worse", "Average", "Better"],
            tickfont=dict(size=11, family="Arial, sans-serif"),
            titlefont=dict(size=12, family="Arial, sans-serif"),
        ),
    )
)

# Horizontal separators between income groups (assuming 4 rows per geography)
fig.add_shape(
    type="line",
    x0=-0.5,
    x1=len(x_labels) - 0.5,
    y0=3.5,
    y1=3.5,
    line=dict(color="white", width=3),
)
fig.add_shape(
    type="line",
    x0=-0.5,
    x1=len(x_labels) - 0.5,
    y0=7.5,
    y1=7.5,
    line=dict(color="white", width=3),
)

fig.update_layout(
    title=dict(
        text="Economic Opportunity Scorecard by Geography and Income Level",
        x=0.5,
        xanchor="center",
        font=dict(size=24, family="Arial, sans-serif", color="#2c3e50"),
    ),
    annotations=[
        dict(
            text=(
                "Mean values across census tracts | "
                "Blue = better outcomes, Red = worse outcomes (normalized across all metrics)"
            ),
            x=0.5,
            y=-0.12,
            xref="paper",
            yref="paper",
            xanchor="center",
            yanchor="top",
            showarrow=False,
            font=dict(size=11, color="#7f8c8d", family="Arial, sans-serif"),
        )
    ],
    xaxis=dict(
        title="",
        side="bottom",
        tickfont=dict(size=11, family="Arial, sans-serif"),
    ),
    yaxis=dict(
        title="",
        tickfont=dict(size=11, family="Arial, sans-serif"),
    ),
    height=700,
    width=1100,
    template="plotly_white",
    margin=dict(t=100, b=100, l=200, r=150),
)

fig.show()

# 6.3 Bar chart comparison of key indicators by income group and geography

import plotly.graph_objects as go
from plotly.subplots import make_subplots

# Metrics to compare and their subplot titles
gap_metrics_subplot = [
    ("Employment Rate", "Employment Rate (%)"),
    ("Bachelors_Plus_Rate", "Bachelor's Degree+ (%)"),
    ("Broadband Rate", "Broadband Access (%)"),
    ("Homeownership Rate", "Homeownership Rate (%)"),
    ("Unemployment Rate", "Unemployment Rate (%)"),
    ("Vacancy Rate", "Vacancy Rate (%)"),
]

# Build comparison data for each metric
subplot_data = {}

for metric, label in gap_metrics_subplot:
    metric_comparison = []

    for geo_name, geo_data in [
        ("Allegheny County", allegheny),
        ("Pennsylvania", pennsylvania),
        ("Nationwide", nationwide),
    ]:
        low_mod = geo_data[geo_data["Distress_Category"] == "Low/Moderate Income"][metric].mean()
        mid_upper = geo_data[geo_data["Distress_Category"] == "Middle/Upper Income"][metric].mean()
        gap = mid_upper - low_mod

        metric_comparison.append(
            {
                "Geography": geo_name,
                "Low/Moderate": low_mod,
                "Middle/Upper": mid_upper,
                "Gap": gap,
            }
        )

    subplot_data[metric] = pd.DataFrame(metric_comparison)

# Create 2x3 grid of subplots
fig = make_subplots(
    rows=2,
    cols=3,
    subplot_titles=[label for _, label in gap_metrics_subplot],
    vertical_spacing=0.15,
    horizontal_spacing=0.10,
)

positions = [(1, 1), (1, 2), (1, 3), (2, 1), (2, 2), (2, 3)]

# Add traces for each metric
for idx, ((metric, label), (row, col)) in enumerate(zip(gap_metrics_subplot, positions)):
    df_metric = subplot_data[metric]

    # Low/Moderate income bars
    fig.add_trace(
        go.Bar(
            name="Low/Moderate Income",
            x=df_metric["Geography"],
            y=df_metric["Low/Moderate"],
            marker_color="#e74c3c",
            text=df_metric["Low/Moderate"].round(1),
            texttemplate="%{text}",
            textposition="outside",
            textfont=dict(size=9),
            showlegend=(idx == 0),
            legendgroup="low_mod",
            hovertemplate="<b>%{x}</b><br>Low/Moderate: %{y:.1f}%<extra></extra>",
        ),
        row=row,
        col=col,
    )

    # Middle/Upper income bars
    fig.add_trace(
        go.Bar(
            name="Middle/Upper Income",
            x=df_metric["Geography"],
            y=df_metric["Middle/Upper"],
            marker_color="#3498db",
            text=df_metric["Middle/Upper"].round(1),
            texttemplate="%{text}",
            textposition="outside",
            textfont=dict(size=9),
            showlegend=(idx == 0),
            legendgroup="mid_upper",
            hovertemplate="<b>%{x}</b><br>Middle/Upper: %{y:.1f}%<extra></extra>",
        ),
        row=row,
        col=col,
    )

    # Axes styling per subplot
    fig.update_yaxes(
        title_text="%",
        title_font=dict(size=10),
        tickfont=dict(size=9),
        gridcolor="#ecf0f1",
        row=row,
        col=col,
    )

    fig.update_xaxes(
        tickfont=dict(size=9),
        tickangle=-45,
        row=row,
        col=col,
    )

# Overall layout
fig.update_layout(
    title=dict(
        text="Opportunity Gaps Across Key Economic Indicators",
        x=0.5,
        xanchor="center",
        font=dict(size=22, family="Arial, sans-serif", color="#2c3e50"),
    ),
    annotations=list(fig.layout.annotations)
    + [
        dict(
            text=(
                "Comparison of Low/Moderate Income vs Middle/Upper Income census tracts "
                "across Allegheny County, Pennsylvania, and the U.S."
            ),
            x=0.5,
            y=-0.08,
            xref="paper",
            yref="paper",
            xanchor="center",
            yanchor="top",
            showarrow=False,
            font=dict(size=11, color="#7f8c8d", family="Arial, sans-serif"),
        )
    ],
    height=800,
    width=1200,
    showlegend=True,
    legend=dict(
        orientation="h",
        yanchor="bottom",
        y=1.05,
        xanchor="center",
        x=0.5,
        font=dict(size=12, family="Arial, sans-serif"),
        bgcolor="rgba(255,255,255,0.8)",
        bordercolor="#bdc3c7",
        borderwidth=1,
    ),
    template="plotly_white",
    barmode="group",
    margin=dict(t=140, b=100, l=60, r=60),
)

fig.show()

# 6.4 Income–employment relationship across geographies

import plotly.express as px
import plotly.graph_objects as go

# Assemble scatter data for all three geographies
scatter_data = []

for geo_name, geo_data in [
    ("Allegheny County", allegheny),
    ("Pennsylvania", pennsylvania),
    ("Nationwide", nationwide),
]:
    temp_df = geo_data[
        [
            "Median Household Income",
            "Employment Rate",
            "FFIEC Tract income level (2022)",
            "Tract Name",
        ]
    ].copy()
    temp_df["Geography"] = geo_name
    scatter_data.append(temp_df)

scatter_df = pd.concat(scatter_data, ignore_index=True)

# Filter out outliers and invalid values
scatter_df = scatter_df[
    (scatter_df["Median Household Income"] > 0)
    & (scatter_df["Median Household Income"] < 250_000)
    & (scatter_df["Employment Rate"] > 0)
    & (scatter_df["Employment Rate"] < 100)
]

# Scatter plot with overall OLS trendline
fig = px.scatter(
    scatter_df,
    x="Median Household Income",
    y="Employment Rate",
    color="Geography",
    color_discrete_map={
        "Allegheny County": "#1f77b4",
        "Pennsylvania": "#9467bd",
        "Nationwide": "#ff7f0e",
    },
    opacity=0.3,
    hover_data={
        "Tract Name": True,
        "FFIEC Tract income level (2022)": True,
        "Median Household Income": ":$,.0f",
        "Employment Rate": ":.1f",
        "Geography": True,
    },
    labels={
        "Median Household Income": "Median Household Income",
        "Employment Rate": "Employment Rate (%)",
    },
    trendline="ols",
    trendline_scope="overall",
    trendline_color_override="#2c3e50",
)

# Emphasize the trendline
for trace in fig.data:
    if getattr(trace, "mode", None) == "lines":
        trace.line.width = 3
        trace.line.dash = "dash"

fig.update_layout(
    title=dict(
        text=(
            "Nation-Level Relationship Between Median Income and Employment Rate"
            "<br><sub>Higher income tracts show stronger employment outcomes</sub>"
        ),
        x=0.5,
        xanchor="center",
        font=dict(size=22, family="Arial, sans-serif", color="#2c3e50"),
    ),
    xaxis=dict(
        title="Median Household Income ($)",
        title_font=dict(size=14, family="Arial, sans-serif"),
        tickformat="$,.0f",
        gridcolor="#ecf0f1",
        tickfont=dict(size=11),
    ),
    yaxis=dict(
        title="Employment Rate (%)",
        title_font=dict(size=14, family="Arial, sans-serif"),
        gridcolor="#ecf0f1",
        range=[70, 100],
        tickfont=dict(size=11),
    ),
    height=600,
    width=1000,
    template="plotly_white",
    plot_bgcolor="white",
    legend=dict(
        title=dict(text="Geography", font=dict(size=13, family="Arial, sans-serif")),
        font=dict(size=12, family="Arial, sans-serif"),
        bgcolor="rgba(255,255,255,0.9)",
        bordercolor="#bdc3c7",
        borderwidth=1,
        x=0.02,
        y=0.98,
        xanchor="left",
        yanchor="top",
    ),
    annotations=[
        dict(
            text="Dashed line shows overall trend across all census tracts",
            x=0.5,
            y=-0.15,
            xref="paper",
            yref="paper",
            xanchor="center",
            yanchor="top",
            showarrow=False,
            font=dict(size=11, color="#7f8c8d", family="Arial, sans-serif"),
        )
    ],
    margin=dict(t=120, b=100, l=80, r=50),
)

fig.show()

# 6.5 Educational attainment by income level and geography

import plotly.graph_objects as go
from plotly.subplots import make_subplots

education_data = []

# Build approximate education breakdown by income level
for geo_name, geo_data in [
    ("Allegheny County", allegheny),
    ("Pennsylvania", pennsylvania),
    ("Nationwide", nationwide),
]:
    for income_level in ["Low", "Moderate", "Middle", "Upper"]:
        subset = geo_data[geo_data["FFIEC Tract income level (2022)"] == income_level]

        if len(subset) == 0:
            continue

        # Bachelor's+ share
        bachelors_plus = subset["Bachelors_Plus_Rate"].mean()

        # Associate degree share relative to population
        associates = (subset["Associates Degree"] / subset["Total Population"] * 100).mean()

        # Approximate "Some College"
        some_college_estimate = associates * 1.5

        # Remainder = high school or less
        hs_or_less = 100 - bachelors_plus - associates - some_college_estimate

        education_data.append(
            {
                "Geography": geo_name,
                "Income Level": income_level,
                "High School or Less": max(0, hs_or_less),
                "Some College": some_college_estimate,
                "Associate Degree": associates,
                "Bachelor's+": bachelors_plus,
            }
        )

education_df = pd.DataFrame(education_data)

# Stacked bars for each geography
fig = make_subplots(
    rows=1,
    cols=3,
    subplot_titles=["Allegheny County", "Pennsylvania", "Nationwide"],
    horizontal_spacing=0.08,
)

colors = {
    "High School or Less": "#e74c3c",
    "Some College": "#e67e22",
    "Associate Degree": "#f39c12",
    "Bachelor's+": "#3498db",
}

education_categories = [
    "High School or Less",
    "Some College",
    "Associate Degree",
    "Bachelor's+",
]

for col_idx, geo_name in enumerate(["Allegheny County", "Pennsylvania", "Nationwide"], 1):
    geo_subset = education_df[education_df["Geography"] == geo_name]

    for edu_cat in education_categories:
        fig.add_trace(
            go.Bar(
                name=edu_cat,
                x=geo_subset["Income Level"],
                y=geo_subset[edu_cat],
                marker_color=colors[edu_cat],
                showlegend=(col_idx == 1),
                legendgroup=edu_cat,
                text=geo_subset[edu_cat].round(1),
                texttemplate="%{text:.0f}%",
                textposition="inside",
                textfont=dict(size=9, color="white"),
                hovertemplate=(
                    "<b>%{x} Income</b><br>"
                    + edu_cat
                    + ": %{y:.1f}%<extra></extra>"
                ),
            ),
            row=1,
            col=col_idx,
        )

    fig.update_xaxes(
        categoryorder="array",
        categoryarray=["Low", "Moderate", "Middle", "Upper"],
        tickfont=dict(size=11),
        row=1,
        col=col_idx,
    )

    fig.update_yaxes(
        title_text="Percentage (%)" if col_idx == 1 else "",
        title_font=dict(size=12),
        tickfont=dict(size=10),
        range=[0, 100],
        gridcolor="#ecf0f1",
        row=1,
        col=col_idx,
    )

fig.update_layout(
    title=dict(
        text="Educational Attainment by Income Level",
        x=0.5,
        xanchor="center",
        font=dict(size=22, family="Arial, sans-serif", color="#2c3e50"),
    ),
    barmode="stack",
    height=550,
    width=1300,
    template="plotly_white",
    legend=dict(
        title=dict(text="Education Level", font=dict(size=13)),
        font=dict(size=12, family="Arial, sans-serif"),
        orientation="h",
        yanchor="bottom",
        y=1.1,
        xanchor="center",
        x=0.5,
        bgcolor="rgba(255,255,255,0.9)",
        bordercolor="#bdc3c7",
        borderwidth=1,
    ),
    annotations=list(fig.layout.annotations)
    + [
        dict(
            text=(
                "Lower-income tracts show higher shares of red/orange (lower attainment), "
                "while upper-income tracts show more blue (Bachelor's+)."
            ),
            x=0.5,
            y=-0.12,
            xref="paper",
            yref="paper",
            xanchor="center",
            yanchor="top",
            showarrow=False,
            font=dict(size=11, color="#7f8c8d"),
        )
    ],
    margin=dict(t=150, b=100, l=70, r=50),
)

fig.show()

print("✓ Education attainment visualization created!")

✓ Education attainment visualization created!

# 6.6 Infrastructure access gaps within Allegheny County

import plotly.graph_objects as go

# Aggregate broadband and transit use by income level (Allegheny only)
infrastructure_data = []

for income_level in ["Low", "Moderate", "Middle", "Upper"]:
    subset = allegheny[allegheny["FFIEC Tract income level (2022)"] == income_level]

    if len(subset) == 0:
        continue

    infrastructure_data.append(
        {
            "Income Level": income_level,
            "Broadband Access": subset["Broadband Rate"].mean(),
            "Public Transit Use": subset["Transit Rate"].mean(),
            "n_tracts": len(subset),
        }
    )

infra_df = pd.DataFrame(infrastructure_data)

# Grouped bar chart: broadband vs transit by income level
fig = go.Figure()

fig.add_trace(
    go.Bar(
        name="Broadband Access",
        x=infra_df["Income Level"],
        y=infra_df["Broadband Access"],
        marker_color="#3498db",
        text=infra_df["Broadband Access"].round(1).astype(str) + "%",
        textposition="outside",
        textfont=dict(size=11),
        hovertemplate=(
            "<b>%{x} Income</b><br>"
            "Broadband Access: %{y:.1f}%<extra></extra>"
        ),
    )
)

fig.add_trace(
    go.Bar(
        name="Public Transit Use",
        x=infra_df["Income Level"],
        y=infra_df["Public Transit Use"],
        marker_color="#e67e22",
        text=infra_df["Public Transit Use"].round(1).astype(str) + "%",
        textposition="outside",
        textfont=dict(size=11),
        hovertemplate=(
            "<b>%{x} Income</b><br>"
            "Public Transit Use: %{y:.1f}%<extra></extra>"
        ),
    )
)

fig.update_layout(
    title=dict(
        text="Infrastructure Access Gaps in Allegheny County",
        x=0.5,
        xanchor="center",
        font=dict(size=22, family="Arial, sans-serif", color="#2c3e50"),
    ),
    xaxis=dict(
        title="Income Level",
        title_font=dict(size=14),
        categoryorder="array",
        categoryarray=["Low", "Moderate", "Middle", "Upper"],
        tickfont=dict(size=12),
    ),
    yaxis=dict(
        title="Percentage (%)",
        title_font=dict(size=14),
        tickfont=dict(size=11),
        gridcolor="#ecf0f1",
        range=[
            0,
            max(
                infra_df["Broadband Access"].max(),
                infra_df["Public Transit Use"].max(),
            )
            + 10,
        ],
    ),
    barmode="group",
    height=550,
    width=900,
    template="plotly_white",
    plot_bgcolor="white",
    legend=dict(
        title=dict(text="Infrastructure Type", font=dict(size=13)),
        font=dict(size=12, family="Arial, sans-serif"),
        orientation="h",
        yanchor="bottom",
        y=1.08,
        xanchor="center",
        x=0.5,
        bgcolor="rgba(255,255,255,0.9)",
        bordercolor="#bdc3c7",
        borderwidth=1,
    ),
    annotations=[
        dict(
            text=(
                "Digital access rises with income, while transit use is highest "
                "in lower-income areas."
            ),
            x=0.5,
            y=-0.15,
            xref="paper",
            yref="paper",
            xanchor="center",
            yanchor="top",
            showarrow=False,
            font=dict(size=11, color="#7f8c8d", family="Arial, sans-serif"),
        )
    ],
    margin=dict(t=130, b=100, l=80, r=50),
)

fig.show()

# Simple numeric summary for the narrative
print("Infrastructure gap summary (Allegheny County):")
broadband_low = infra_df.loc[infra_df["Income Level"] == "Low", "Broadband Access"].values[0]
broadband_upper = infra_df.loc[infra_df["Income Level"] == "Upper", "Broadband Access"].values[0]
transit_low = infra_df.loc[infra_df["Income Level"] == "Low", "Public Transit Use"].values[0]
transit_upper = infra_df.loc[infra_df["Income Level"] == "Upper", "Public Transit Use"].values[0]

print(f"- Broadband gap (Upper - Low): {broadband_upper - broadband_low:.1f} percentage points")
print(f"- Transit use: higher in {'lower' if transit_low > transit_upper else 'upper'}-income areas")

Infrastructure gap summary (Allegheny County):
- Broadband gap (Upper - Low): 5.8 percentage points
- Transit use: higher in lower-income areas

# 6.7 Choropleth: income classification by census tract (Allegheny County)

import geopandas as gpd
import plotly.express as px
import plotly.graph_objects as go

# Path to Pennsylvania tract shapefile (2022)
shapefile_path = "/Users/sofiahutton/Documents/Fall 2025 CMU Classes/visualizations with python /tl_2022_42_tract.shp"

# Load tracts and filter to Allegheny County (county FIPS = 003)
gdf = gpd.read_file(shapefile_path)
gdf_allegheny = gdf[gdf["COUNTYFP"] == "003"].copy()

# Build GEOID for merge
gdf_allegheny["GEOID_full"] = gdf_allegheny["GEOID"]

allegheny_for_map = allegheny.copy()
allegheny_for_map["GEOID_full"] = (
    allegheny_for_map["state"].astype(str).str.zfill(2)
    + allegheny_for_map["county"].astype(str).str.zfill(3)
    + allegheny_for_map["Tract Code (6-digit)"].astype(str).str.zfill(6)
)

# Merge shapefile with tract-level indicators
gdf_merged = gdf_allegheny.merge(
    allegheny_for_map[
        [
            "GEOID_full",
            "FFIEC Tract income level (2022)",
            "Tract Name",
            "Employment Rate",
            "Median Household Income",
            "Bachelors_Plus_Rate",
        ]
    ],
    on="GEOID_full",
    how="left",
)

# Income-level colors
color_map = {
    "Low": "#e74c3c",
    "Moderate": "#e67e22",
    "Middle": "#f39c12",
    "Upper": "#3498db",
    "Unknown": "#95a5a6",
}

# Fill missing classifications
gdf_merged["FFIEC Tract income level (2022)"] = gdf_merged[
    "FFIEC Tract income level (2022)"
].fillna("Unknown")

# Center map around tract centroids
center_lat = gdf_merged.geometry.centroid.y.mean()
center_lon = gdf_merged.geometry.centroid.x.mean()

fig = px.choropleth_mapbox(
    gdf_merged,
    geojson=gdf_merged.geometry,
    locations=gdf_merged.index,
    color="FFIEC Tract income level (2022)",
    color_discrete_map=color_map,
    category_orders={
        "FFIEC Tract income level (2022)": ["Low", "Moderate", "Middle", "Upper", "Unknown"]
    },
    mapbox_style="carto-positron",
    center={"lat": center_lat, "lon": center_lon},
    zoom=9,
    opacity=0.7,
    hover_data={
        "Tract Name": True,
        "FFIEC Tract income level (2022)": True,
        "Employment Rate": ":.1f",
        "Median Household Income": ":$,.0f",
        "Bachelors_Plus_Rate": ":.1f",
    },
    labels={
        "FFIEC Tract income level (2022)": "Income Level",
        "Employment Rate": "Employment Rate (%)",
        "Median Household Income": "Median Income",
        "Bachelors_Plus_Rate": "Bachelor's Degree+ (%)",
    },
)

fig.update_layout(
    title=dict(
        text="Income Classification by Census Tract<br><sub>Allegheny County, Pennsylvania</sub>",
        x=0.5,
        xanchor="center",
        font=dict(size=24, family="Arial, sans-serif", color="#2c3e50"),
    ),
    margin=dict(r=0, t=80, l=0, b=0),
    height=700,
    legend=dict(
        title=dict(text="FFIEC Income Level", font=dict(size=14)),
        font=dict(size=12, family="Arial, sans-serif"),
        bgcolor="rgba(255,255,255,0.9)",
        bordercolor="#bdc3c7",
        borderwidth=2,
        x=0.02,
        y=0.98,
        xanchor="left",
        yanchor="top",
    ),
)

fig.show()

# 6.8 Compound disadvantage hot spots in Allegheny County

import geopandas as gpd
import plotly.express as px
import numpy as np

# Load Pennsylvania tract shapefile and filter to Allegheny County
shapefile_path = "/Users/sofiahutton/Documents/Fall 2025 CMU Classes/visualizations with python /tl_2022_42_tract.shp"

gdf = gpd.read_file(shapefile_path)
gdf_allegheny = gdf[gdf["COUNTYFP"] == "003"].copy()

# Create compound disadvantage score for Allegheny tracts
allegheny_hotspot = allegheny.copy()

# Normalize component metrics to a 0–1 scale where higher = worse outcome
allegheny_hotspot["unemployment_score"] = allegheny_hotspot["Unemployment Rate"] / 100
allegheny_hotspot["vacancy_score"] = allegheny_hotspot["Vacancy Rate"] / 100
allegheny_hotspot["low_employment_score"] = (100 - allegheny_hotspot["Employment Rate"]) / 100
allegheny_hotspot["low_education_score"] = (100 - allegheny_hotspot["Bachelors_Plus_Rate"]) / 100
allegheny_hotspot["low_broadband_score"] = (100 - allegheny_hotspot["Broadband Rate"]) / 100
allegheny_hotspot["low_homeownership_score"] = (100 - allegheny_hotspot["Homeownership Rate"]) / 100

disadvantage_components = [
    "unemployment_score",
    "vacancy_score",
    "low_employment_score",
    "low_education_score",
    "low_broadband_score",
    "low_homeownership_score",
]

# Average across components and rescale to 0–100
allegheny_hotspot["Disadvantage_Score"] = allegheny_hotspot[disadvantage_components].mean(axis=1) * 100

# Distribution check
print("Disadvantage score distribution (Allegheny tracts):")
print(f"- Min:              {allegheny_hotspot['Disadvantage_Score'].min():.1f}")
print(f"- 25th percentile:  {allegheny_hotspot['Disadvantage_Score'].quantile(0.25):.1f}")
print(f"- Median:           {allegheny_hotspot['Disadvantage_Score'].median():.1f}")
print(f"- 75th percentile:  {allegheny_hotspot['Disadvantage_Score'].quantile(0.75):.1f}")
print(f"- Max:              {allegheny_hotspot['Disadvantage_Score'].max():.1f}")

# Quartile categories for clearer visual breakpoints
allegheny_hotspot["Disadvantage_Quartile"] = pd.qcut(
    allegheny_hotspot["Disadvantage_Score"],
    q=4,
    labels=["Lowest 25%", "Lower-Middle 25%", "Upper-Middle 25%", "Highest 25%"],
)

print("\nQuartile counts:")
print(allegheny_hotspot["Disadvantage_Quartile"].value_counts().sort_index())

# Merge with Allegheny tracts shapefile
gdf_allegheny["GEOID_full"] = gdf_allegheny["GEOID"]
allegheny_hotspot["GEOID_full"] = (
    allegheny_hotspot["state"].astype(str).str.zfill(2)
    + allegheny_hotspot["county"].astype(str).str.zfill(3)
    + allegheny_hotspot["Tract Code (6-digit)"].astype(str).str.zfill(6)
)

gdf_hotspot = gdf_allegheny.merge(
    allegheny_hotspot[
        [
            "GEOID_full",
            "Disadvantage_Score",
            "Disadvantage_Quartile",
            "Tract Name",
            "Employment Rate",
            "Unemployment Rate",
            "Bachelors_Plus_Rate",
            "Broadband Rate",
            "Homeownership Rate",
            "FFIEC Tract income level (2022)",
        ]
    ],
    on="GEOID_full",
    how="left",
)

# Discrete color scale by quartile (green = best, dark red = worst)
color_map_discrete = {
    "Lowest 25%": "#2ecc71",        # Green (least disadvantaged)
    "Lower-Middle 25%": "#f39c12",  # Yellow
    "Upper-Middle 25%": "#e67e22",  # Orange
    "Highest 25%": "#c0392b",       # Dark red (most disadvantaged)
}

fig = px.choropleth_mapbox(
    gdf_hotspot,
    geojson=gdf_hotspot.geometry,
    locations=gdf_hotspot.index,
    color="Disadvantage_Quartile",
    color_discrete_map=color_map_discrete,
    category_orders={
        "Disadvantage_Quartile": [
            "Lowest 25%",
            "Lower-Middle 25%",
            "Upper-Middle 25%",
            "Highest 25%",
        ]
    },
    mapbox_style="carto-positron",
    center={
        "lat": gdf_hotspot.geometry.centroid.y.mean(),
        "lon": gdf_hotspot.geometry.centroid.x.mean(),
    },
    zoom=9,
    opacity=0.8,
    hover_data={
        "Tract Name": True,
        "Disadvantage_Score": ":.1f",
        "Disadvantage_Quartile": True,
        "FFIEC Tract income level (2022)": True,
        "Employment Rate": ":.1f",
        "Unemployment Rate": ":.1f",
        "Bachelors_Plus_Rate": ":.1f",
        "Broadband Rate": ":.1f",
        "Homeownership Rate": ":.1f",
    },
    labels={
        "Disadvantage_Quartile": "Disadvantage Level",
        "Disadvantage_Score": "Composite Score",
        "FFIEC Tract income level (2022)": "Income Level",
        "Employment Rate": "Employment (%)",
        "Unemployment Rate": "Unemployment (%)",
        "Bachelors_Plus_Rate": "Bachelor's+ (%)",
        "Broadband Rate": "Broadband (%)",
        "Homeownership Rate": "Homeownership (%)",
    },
)

fig.update_layout(
    title=dict(
        text=(
            "Compound Disadvantage Hot Spots in Allegheny County"
            "<br><sub>Dark red = top 25% most disadvantaged tracts (multiple overlapping challenges)</sub>"
        ),
        x=0.5,
        xanchor="center",
        font=dict(size=22, family="Arial, sans-serif", color="#2c3e50"),
    ),
    margin=dict(r=0, t=100, l=0, b=0),
    height=700,
    legend=dict(
        title=dict(text="Disadvantage Level", font=dict(size=13)),
        font=dict(size=12, family="Arial, sans-serif"),
        bgcolor="rgba(255,255,255,0.95)",
        bordercolor="#bdc3c7",
        borderwidth=2,
        x=0.02,
        y=0.98,
        xanchor="left",
        yanchor="top",
    ),
)

fig.show()

print("\nHot spot map generated: dark red tracts represent the highest composite disadvantage; green tracts the lowest.")

Disadvantage score distribution (Allegheny tracts):
- Min:              18.4
- 25th percentile:  29.0
- Median:           32.4
- 75th percentile:  36.7
- Max:              100.0

Quartile counts:
Disadvantage_Quartile
Lowest 25%          97
Lower-Middle 25%    96
Upper-Middle 25%    96
Highest 25%         97
Name: count, dtype: int64

Hot spot map generated: dark red tracts represent the highest composite disadvantage; green tracts the lowest.

# 6.9 "Success story" tracts: low/moderate-income areas with strong outcomes

import plotly.express as px
import plotly.graph_objects as go

# Start from low/moderate-income tracts only
allegheny_success = allegheny[allegheny["Distress_Category"] == "Low/Moderate Income"].copy()

# Benchmark employment for low/moderate tracts
low_mod_median_employment = allegheny_success["Employment Rate"].median()
low_mod_75th_employment = allegheny_success["Employment Rate"].quantile(0.75)

# Middle/upper-income benchmark
mid_upper_avg_employment = allegheny[
    allegheny["Distress_Category"] == "Middle/Upper Income"
]["Employment Rate"].mean()

print("Employment benchmarks (Allegheny County):")
print("Low/Moderate income tracts:")
print(f"- Median employment rate:     {low_mod_median_employment:.1f}%")
print(f"- 75th percentile:           {low_mod_75th_employment:.1f}%")
print("\nMiddle/Upper income tracts:")
print(f"- Average employment rate:   {mid_upper_avg_employment:.1f}%")

# Flag low/moderate tracts that match or exceed middle/upper employment
allegheny_success["Success_Story"] = (
    allegheny_success["Employment Rate"] >= mid_upper_avg_employment
)

# Benchmark for education
mid_upper_avg_education = allegheny[
    allegheny["Distress_Category"] == "Middle/Upper Income"
]["Bachelors_Plus_Rate"].mean()
allegheny_success["High_Education"] = (
    allegheny_success["Bachelors_Plus_Rate"] >= mid_upper_avg_education
)

# Categorize performance within low/moderate tracts
def categorize_success(row):
    if row["Success_Story"] and row["High_Education"]:
        return "Excelling (High Employment + Education)"
    elif row["Success_Story"]:
        return "Strong Employment"
    elif row["High_Education"]:
        return "Strong Education"
    else:
        return "Struggling"

allegheny_success["Performance_Category"] = allegheny_success.apply(
    categorize_success, axis=1
)

print("\nPerformance category counts (Low/Moderate income tracts):")
print(allegheny_success["Performance_Category"].value_counts())

# Scatter plot: success profiles among low/moderate tracts
fig = px.scatter(
    allegheny_success,
    x="Median Household Income",
    y="Employment Rate",
    size="Bachelors_Plus_Rate",
    color="Performance_Category",
    color_discrete_map={
        "Excelling (High Employment + Education)": "#27ae60",  # Green
        "Strong Employment": "#3498db",                        # Blue
        "Strong Education": "#9b59b6",                         # Purple
        "Struggling": "#e74c3c",                               # Red
    },
    hover_data={
        "Tract Name": True,
        "FFIEC Tract income level (2022)": True,
        "Employment Rate": ":.1f",
        "Bachelors_Plus_Rate": ":.1f",
        "Median Household Income": ":$,.0f",
        "Broadband Rate": ":.1f",
        "Homeownership Rate": ":.1f",
    },
    labels={
        "Employment Rate": "Employment Rate (%)",
        "Median Household Income": "Median Household Income",
        "Bachelors_Plus_Rate": "Bachelor's+ (%)",
        "Performance_Category": "Performance",
    },
    size_max=20,
)

# Reference line: middle/upper-income average employment
fig.add_hline(
    y=mid_upper_avg_employment,
    line_dash="dash",
    line_color="#95a5a6",
    annotation_text=f"Middle/Upper avg employment ({mid_upper_avg_employment:.1f}%)",
    annotation_position="right",
)

fig.update_layout(
    title=dict(
        text=(
            '"Success Stories": Low/Moderate Income Tracts with Strong Outcomes'
            "<br><sub>Green/blue points mark tracts matching or exceeding middle/upper-income benchmarks</sub>"
        ),
        x=0.5,
        xanchor="center",
        font=dict(size=20, family="Arial, sans-serif", color="#2c3e50"),
    ),
    xaxis=dict(
        title="Median Household Income ($)",
        title_font=dict(size=14),
        tickformat="$,.0f",
        gridcolor="#ecf0f1",
        tickfont=dict(size=11),
    ),
    yaxis=dict(
        title="Employment Rate (%)",
        title_font=dict(size=14),
        gridcolor="#ecf0f1",
        tickfont=dict(size=11),
        range=[75, 100],
    ),
    height=650,
    width=1100,
    template="plotly_white",
    plot_bgcolor="white",
    legend=dict(
        title=dict(text="Tract Performance", font=dict(size=13)),
        font=dict(size=11, family="Arial, sans-serif"),
        bgcolor="rgba(255,255,255,0.95)",
        bordercolor="#bdc3c7",
        borderwidth=1,
        x=0.02,
        y=0.98,
        xanchor="left",
        yanchor="top",
    ),
    annotations=list(fig.layout.annotations)
    + [
        dict(
            text=(
                "Bubble size = Bachelor's+ share. "
                "Dashed line shows the middle/upper-income average employment rate."
            ),
            x=0.5,
            y=-0.12,
            xref="paper",
            yref="paper",
            xanchor="center",
            yanchor="top",
            showarrow=False,
            font=dict(size=11, color="#7f8c8d", family="Arial, sans-serif"),
        )
    ],
    margin=dict(t=130, b=100, l=80, r=50),
)

fig.show()

# Simple numeric summary
n_success = allegheny_success["Success_Story"].sum()
n_total = len(allegheny_success)

print(f"\nSummary: {n_success} out of {n_total} low/moderate-income tracts")
print("have employment rates at or above the middle/upper-income average.")
print("These tracts may offer useful lessons about local conditions and interventions.")

Employment benchmarks (Allegheny County):
Low/Moderate income tracts:
- Median employment rate:     92.3%
- 75th percentile:           95.1%

Middle/Upper income tracts:
- Average employment rate:   95.8%

Performance category counts (Low/Moderate income tracts):
Performance_Category
Struggling                                 97
Strong Employment                          16
Strong Education                            4
Excelling (High Employment + Education)     4
Name: count, dtype: int64

Summary: 20 out of 121 low/moderate-income tracts
have employment rates at or above the middle/upper-income average.
These tracts may offer useful lessons about local conditions and interventions.

	NAME	B23025_002E	B23025_003E	B23025_004E	B23025_005E	B15003_021E	B15003_022E	B25003_001E	B25003_002E	B25070_007E	...	B25002_003E	B19013_001E	B28002_002E	B08301_010E	B01003_001E	state	county	tract	tract_6	geoid
0	Census Tract 201; Autauga County; Alabama	738	732	713	19	84	182	700	519	0	...	33	60563	643	0	1865	01	001	020100	020100	01001020100
1	Census Tract 202; Autauga County; Alabama	947	919	868	51	86	163	544	429	8	...	136	57460	427	0	1861	01	001	020200	020200	01001020200
2	Census Tract 203; Autauga County; Alabama	1808	1781	1748	33	284	209	1305	912	9	...	126	77371	1170	0	3492	01	001	020300	020300	01001020300
3	Census Tract 204; Autauga County; Alabama	1875	1854	1837	17	302	662	1666	1306	0	...	56	73191	1563	53	3987	01	001	020400	020400	01001020400
4	Census Tract 205.01; Autauga County; Alabama	2504	2400	2386	14	319	763	1783	971	21	...	74	79953	1759	72	4121	01	001	020501	020501	01001020501

	NAME	B23025_002E	B23025_003E	B23025_004E	B23025_005E	B15003_021E	B15003_022E	B25003_001E	B25003_002E	B25070_007E	...	Tract MFI	Tract income percentage	Tract income level	employment_rate	unemployment_rate	homeownership_rate	vacancy_rate	broadband_rate	transit_rate	rent_burdened_count
0	Census Tract 201; Autauga County; Alabama	738	732	713	19	84	182	700	519	0	...	68115.0	103.79	Middle	0.974044	0.025956	0.741429	0.045020	0.344772	0.000000	74
1	Census Tract 202; Autauga County; Alabama	947	919	868	51	86	163	544	429	8	...	68115.0	73.60	Moderate	0.944505	0.055495	0.788603	0.200000	0.229447	0.000000	66
2	Census Tract 203; Autauga County; Alabama	1808	1781	1748	33	284	209	1305	912	9	...	68115.0	102.93	Middle	0.981471	0.018529	0.698851	0.088050	0.335052	0.000000	190
3	Census Tract 204; Autauga County; Alabama	1875	1854	1837	17	302	662	1666	1306	0	...	68115.0	110.95	Middle	0.990831	0.009169	0.783914	0.032520	0.392024	0.013293	8
4	Census Tract 205.01; Autauga County; Alabama	2504	2400	2386	14	319	763	1783	971	21	...	68115.0	133.41	Upper	0.994167	0.005833	0.544588	0.039849	0.426838	0.017471	398

	Tract Name	Labor Force	Employed	Unemployed	Associates Degree	Bachelors or Higher	Total Housing Units	Owner-Occupied	Vacant Units	Median Household Income	...	Employment Rate	Unemployment Rate	Homeownership Rate	Vacancy Rate	Broadband Rate	Transit Rate	Rent Burdened (30%+)	Distress_Category	Bachelors_Plus_Rate	Rent_Burden_Rate
0	020100	732	713	19	84	182	700	519	33	60563	...	97.404372	2.595628	74.142857	4.502046	34.477212	0.000000	74	Middle/Upper Income	9.758713	10.571429
1	020200	919	868	51	86	163	544	429	136	57460	...	94.450490	5.549510	78.860294	20.000000	22.944653	0.000000	66	Low/Moderate Income	8.758732	12.132353
2	020300	1781	1748	33	284	209	1305	912	126	77371	...	98.147108	1.852892	69.885057	8.805031	33.505155	0.000000	190	Middle/Upper Income	5.985109	14.559387
3	020400	1854	1837	17	302	662	1666	1306	56	73191	...	99.083064	0.916936	78.391357	3.252033	39.202408	1.329320	8	Middle/Upper Income	16.603963	0.480192
4	020501	2400	2386	14	319	763	1783	971	74	79953	...	99.416667	0.583333	54.458777	3.984922	42.683815	1.747149	398	Middle/Upper Income	18.514924	22.321929

Economic Opportunity Gaps in Allegheny County: A Tract-Level Analysis¶

EDA/Data Visualization (90-800) Final Project¶

Sofia Hutton | Sdhutton | December 2025¶

AI Disclaimer¶

Abstract¶

1. Introduction & Motivation¶

Why This Project?¶

Personal Connection¶

Policy Implications¶

Research Questions¶

2. Dataset Description¶

Data Sources¶

2.1 American Community Survey (ACS) 5-Year Estimates (2018-2022)¶

2.2 FFIEC Income Classification Data (2022)¶

2.3 TIGER/Line Shapefiles - Pennsylvania Census Tracts (2022)¶

2.4 (Inspiration) ICIC and SRI's Economic Connectivity Dashboard¶

3. Initial Hypotheses¶

4. Data Collection & Processing¶

4.1 System and Census API Configuration¶

4.2 Retrieve Census Data¶

4.3 Build Skeleton Dataframe¶

4.4 Merge FFIEC Income Data¶

4.5 Calculate Derived Metrics¶

4.6 Create Clean Output Table¶

5. Exploratory Data Analysis¶

5.1 Summary Statistics -¶

5.1 Key Patterns Observed¶

5.2 Income Distribution Profiles: Allegheny vs. Pennsylvania vs. Nationwide¶

5.2 Key Patterns: Income Distribution Profiles¶

5.3 Opportunity Gaps: Lower-Income vs. Higher-Income Tracts¶

5.3 Key Patterns: Opportunity Gaps Between Lower-Income and Higher-Income Tracts¶

6. Visualizations¶

6.1 Opportunity Gaps Across Income Levels¶

6.2 Economic Opportunity Scorecard by Geography and Income Level¶

6.3 Opportunity Gaps Across Key Economic Indicators¶

6.4 Relationship Between Median Income and Employment Rate¶

6.5 Educational Attainment by Income Level¶

6.6 Infrastructure Access Gaps in Allegheny County¶

6.7 Income Classification by Census Tract - Allegheny County¶

6.8 Compound Disadvantage Hot Spots in Allegheny County¶

6.9 "Success Stories" - Low/Moderate Income Tracts Beating the Odds¶

7. Key Findings¶

Finding 1: Unemployment Gaps Are Severe and Persistent¶

Finding 2: The Education Divide Is a Bachelor's Degree Cliff¶

Finding 3: Disadvantage Is Multi-Dimensional, Not Isolated¶

Finding 4: Allegheny's Unique Challenges Are Broadband and Housing Vacancy¶

Finding 5: Disadvantage Clusters Geographically—It's Not Random¶

Finding 6: Transit Dependency Reveals Constrained Mobility, Not Preference¶

Finding 7: Success Stories Exist—Disadvantage Isn't Destiny¶

8. Conclusions & Policy Recommendations¶

What Did This Analysis Prove?¶

Policy Recommendations¶

Limitations¶

Future Research¶

9. References¶